Lipid metabolism of Acetobacter pasteurianus and its main components with hypoglycemic effects

Abstract

BACKGROUND

Probiotic Acetobacter pasteurianus is used to treat diabetes, but its specific hypoglycemic substances and mechanisms remain unclear.

AIM

To investigate the components for lipid metabolism of A. pasteurianus and its hypoglycemic effects, providing a basis for its broader application.

METHODS

The lipid metabolism of A. pasteurianus under different growth conditions was analyzed using lipidomics. Neutral lipid staining in A. pasteurianus cells and the formation of lipid droplet-like structures were observed using a confocal laser scanning microscope. The neutral lipid components were also analyzed using thin layer chromatography. A diabetic mouse model was established to evaluate the hypoglycemic effects of the main lipid components of A. pasteurianus and their role in repairing tissues such as the pancreas.

RESULTS

After comparing the effects of three culture media, namely, brain heart infusion (BHI) medium with 2% glucose, chromium-rich and zinc-rich medium, and mineral salt medium, A. pasteurianus grew well in BHI containing 2% glucose and produced the most lipids. A total of 583 lipid metabolic products was identified, with higher levels of coenzyme Q9 (CoQ9), oleic acid (OA), and wax ester, but no triacylglycerol was observed. It was found that the components that affected lipid metabolism in A. pasteurianus were mainly CoQ9 and OA. They exhibited hypoglycemic effects comparable to metformin in diabetic mice, repaired damaged pancreatic tissues, and did not cause damage to the liver and spleen.

CONCLUSION

Under high-nutrient growth conditions, A. pasteurianus contains abundant lipid components, such as CoQ9 and OA, with good hypoglycemic effects.

Key Words: Acetobacter pasteurianus; Lipid metabolism; Neutral lipids; Lipid droplets; Hypoglycemic effects

Core Tip: Research has shown that Acetobacter pasteurianus plays an effective role in alleviating diabetes. Therefore, we investigated its lipid metabolism, main lipid components, and hypoglycemic effects. We found that A. pasteurianus produces various lipids, with coenzyme Q9 and oleic acid being the most abundant components that have hypoglycemic effects, and can repair damaged pancreatic tissues. This provides a theoretical basis for promoting A. pasteurianus for its antidiabetic properties.

INTRODUCTION

The Global Burden of Disease Study (2021), published online in The Lancet in 2023, reports that there were 529 million diabetic patients worldwide in 2021, and this figure is projected to reach 1.31 billion globally that by 2050[1]. Probiotics are a group of beneficial live microorganisms that can regulate blood glucose, blood lipids, cholesterol levels, and improve insulin sensitivity[2,3]. Acetobacter is a bacteria that oxidizes ethanol to acetic acid (AA) under the action of alcohol dehydrogenases and acetaldehyde dehydrogenase[4]. Therefore, it is widely applied as a probiotic in the fermentation industry for producing AA, gluconic acid, and sorbitol[5]. Acetobacter pasteurianus, a key component of kombucha[6], plays an adjunctive role in treating hypertension, hyperlipidemia, and hyperglycemia[7]. A. pasteurianus, containing nicotinamide adenine dinucleotide coenzyme and glucuronate dehydrogenase, demonstrates good effects on reducing blood glucose and alleviating diabetes[8].

Some safe probiotics, such as Lactobacillus and Bifidobacterium that have functions of anti-inflammation, regulating intestinal flora, or promoting the production of short-chain fatty acids (SCFAs), can improve hyperlipidemia, hypertension, and diabetes[3,9]. Especially SCFAs can regulate the immune system, peripheral tissue energy metabolism, insulin sensitivity, and blood lipid levels[10]. A. pasteurianus can play an important role in the treatment of diabetes by regulating intestinal flora and metabolites[11], self-generation, or fermentation of dietary fiber SCFAs (e.g., AA)[12,13], and can decompose the substrate glucose, serving as a carbon source for growth. Other probiotics are not as effective as A. pasteurianus. So, exploring the beneficial components synthesized from A. pasteurianus can help us understand hyperglycemic metabolic regulation, which is crucial for developing better treatments for diabetes.

Additionally, lipid droplets (LDs) are emerging organelles, consisting of a single phospholipid membrane, a neutral lipid core, and various endoplasmic reticulum (ER)-resident proteins[14-16]. In addition to the common triacylglycerol (TAG), the neutral lipids in LDs also include cholesteryl ester, stanol ester, and wax ester (WE). Apart from eukaryotes, a significant amount of neutral lipids also accumulates as the core of LDs in the cytoplasm of bacteria, such as probiotics[17,18]. LDs play a key role in the synthesis, storage, transport, modification, and hydrolysis of lipids[19]. Excessive lipid storage, particularly the accumulation of ectopic lipids, is a major risk factor for metabolic syndrome disorders such as obesity, fatty liver, atherosclerosis, and diabetes[20].

Therefore, studying lipids is crucial for treating metabolic diseases. However, there are no reports on the lipid metabolic components and bioactivities of A. pasteurianus. Thus, this study conducted a preliminary exploration around its lipid metabolism and hypoglycemic function, identifying the main lipid components with hypoglycemic effects, to provide a theoretical basis for its clinical application value.

MATERIALS AND METHODS

Experimental reagents and strains

The experimental reagents used were as follows: glucose (20200403; Guangdong Guanghua Sci-Tech Co., Ltd., Shantou, China); yeast extract (20200422; Beijing Aoboxing Bio-Tech Co., Ltd., Beijing, China); agar (310C022; Beijing Solarbio Science and Technology Co., Ltd.); brain heart infusion (BHI) medium (Lot No. 3555372; OXOID, Basingstoke, Hampshire, United Kingdom); anhydrous ethanol (C11974944; Shanghai Macklin Biochemical Co., Ltd., Shanghai, China); calcium carbonate (CaCO₃) (P1260108; Titan Technology Shanghai Co., Ltd., Shanghai, China); HCS LipidTOX™ Red Neutral Lipid Stain (H34476; Invitrogen, Carlsbad, CA, United States); lipid standards (Sigma-Aldrich, St. Louis, MO, United States); sodium chloride (NaCl) (S7653), potassium chloride (P9333), magnesium chloride (MgCl₂) (208337), potassium dihydrogen phosphate (P5655), disodium hydrogen phosphate (S5136), sucrose (S9378), tricine (T0377), chloroform (CHCl₃) (C2432), acetone (650501), n-Hexane (296090), diethyl ether (C₂H₅)₂O (296082), AA (320099), methanol (MeOH) (322415), glycerol (G5516), and iodine (207772) purchased from Sigma-Aldrich; poly-L-lysine (P8920; Sigma); anti-fade reagent (P0126; Beyotime Biotech Inc., Beijing, China); streptozotocin (STZ) (C20PA038100B), citric acid (CA) (C10723907), and sodium citrate (C10712912) purchased from Shanghai Macklin Biochemical Co., Ltd. (Shanghai, China); and metformin (Met) (2004032; Beijing Jingfeng Pharmaceutical Group Co., Ltd., Beijing, China). A. pasteurianus strains (obtained from Guangdong Microbial Culture Collection Center (GIM1. 67; Guangzhou, China) were used.

Bacterial media and buffers

The bacterial media and buffers used in the experiments were as follows: (1) Liquid medium composed of 10 g/L glucose, 10 g/L yeast extract, and 20 mL/L anhydrous ethanol; (2) Solid medium composed of the aforementioned liquid medium treated with 15 g/L CaCO₃ and 20 g/L agar; and (3) Mineral salt medium (MSM) composed of 0.5 g/L ammonium chloride as a nitrogen source (low-content nitrogen), 10 g/L sodium gluconate (E576) as a carbon source (high-content carbon), 0.2 g/L magnesium sulfate heptahydrate, 0.02 g/L calcium chloride dihydrate, 1.5 g/L potassium dihydrogen phosphate, 0.0012 g/L ferric ammonium citrate, 9 g/L disodium hydrogen phosphate dodecahydrate, 0.1 mL SL6 (composition: 1.0 g/L zinc sulfate heptahydrate, 0.3 g/L manganese (II) chloride tetrahydrate, 3.0 g/L boric acid, 2.0 g/L cobalt (II) chloride hexahydrate, 0.1 g/L copper (II) chloride dihydrate, 0.2 g/L nickel (II) chloride hexahydrate, and 0.3 g/L sodium molybdate (VI) dihydrate). (4) Buffer A: 25 mmol/L tricine, 250 mmol/L sucrose, and 6 mmol/L potassium hydroxide (KOH) (potential of hydrogen [pH] = 7.8). (5) Buffer B: 20 mmol/L 4-(2-hydroxyerhyl)piperazine-1-erhanesulfonic acid, 100 mmol/L NaCl, 2 mmol/L MgCl₂, and 12 mmol/L KOH (pH = 7.4).

Metabolomics analysis

Preparation of A. pasteurianus samples: (1) A. pasteurianus cells were resuscitated, cultured in a solution or BHI until they reached the logarithmic stage; (2) The bacterial solution, with its initial concentration adjusted to 1 × 10⁴ culture-forming units/mL, was treated with chromium trichloride (64 μg/mL) and zinc chloride, cultured for 48 hours, collected after it had been cultured again, and centrifuged to remove the supernatant. Thereafter, the precipitate was washed using phosphate-buffered saline (PBS) to obtain A. pasteurianus cells rich in chromium and zinc; (3) The bacterial solution prepared as discussed in (1) was centrifuged. The precipitate was transferred to a conical flask, cultured in an appropriate amount of MSM for 3 to 4 days, centrifuged, and collected; and (4) Three samples were obtained from each of the three A. pasteurianus precipitates prepared above (cultured in BHI, rich in chromium and zinc, and cultured in MSM), labeled as B1-B3, F1-F3, and M1-M3, respectively, and sent to Beijing Novogene Bioinformatics Technology Co., Ltd. (Beijing, China) for analysis.

Pretreatment of samples: (1) The sample (20 mg) was thawed on ice and weighed, added to the lipid extraction solution (1 mL) (methyl tert-butyl ether = 3:1, V/V), and shaken at 2500 rpm at room temperature for 15 minutes; (2) The sample was centrifuged at 12000 rpm for 10 minutes at 4 °C. Thereafter, the supernatant (500 μL) was transferred to a correspondingly numbered Eppendorf (EP) tube and concentrated at 20 °C until completely dry; and (3) The sample was treated with a lipid resuspension solution (200 μL) (acetonitrile [C₂H₃N] = 1:1, V/V), vortexed for 3 minutes, and centrifuged at 12000 rpm for 3 minutes, with the supernatant transferred for liquid chromatography-tandem mass spectrometry (LC-MS/MS) analysis.

Conditions of acquiring data using chromatography-MS: The data acquisition system primarily included ultra-performance LC (ExionLC™ AD; SCIEX, Framingham, MA, United States) and MS/MS (QTRAP^® 6500+ System; SCIEX). The conditions of acquiring data using LC mainly included: (1) Column: Thermo Accucore™ C30 column (2.6 μm, 2.1 mm × 100 mm; Thermo Fisher Scientific, Waltham, MA, United States); mobile phases composed of formic acid (0.1%) and ammonium formate (10 mmol/L): Phase A, C₂H₃N/water (60/40, V/V), and Phase B, C₂H₃N/isopropanol (10/90, V/V); (2) Gradient elution: At 0 minute, the ratio of A/B was 80:20 (V/V); At 2 minutes, it changed to 70:30 (V/V); At 4 minutes, it was adjusted to 40:60 (V/V); At 9 minutes, it shifted to 15:85 (V/V); At 14 minutes, it was 10:90 (V/V); At 15.5 minutes, it reached 5:95 (V/V); At 17.3 minutes, it remained 5:95 (V/V); At 17.5 minutes, it returned to 80:20 (V/V); At 20 minutes, it remained at 80:20 (V/V); and (3) Flow rate of 0.35 mL/minute. The column temperature was 45 °C. The injection volume was 2 μL.

Qualitative and quantitative analyses of lipids: The qualitative analysis was performed based on the self-built MetWare Database, using the retention time and the ion pair information of the detected substances. Lipid quantification was completed using the multiple reaction monitoring mode of a triple quadrupole mass spectrometer (Supplementary Figure 1). After elimination of non-target ion interferences, chromatographic peak integrations of different lipid peak areas were performed and quantitatively analyzed using the internal standard method. The ratio of the integrated peak area was substituted into a linear equation and calculation formula to determine the content of the substance in the sample. The calculation formula was X = R × c × F × V/m, where X: Lipid content in the sample (pmol/g); R: Ratio of peak area of the analyte to the internal standard; F: Correction factor for different types of internal standards; c: Internal standard concentration (μmol/L); V: Volume of the sample extract (μL); and m: Mass of the sample taken (g).

Neutral lipid staining

(1) The bacterial solution (1 mL) was centrifuged to remove the supernatant and washed once using PBS before being resuspended; (2) A total of 150 μL of resuspended bacterial solution was placed onto a coverslip pretreated with poly-L-lysine, and the excess bacterial solution was rinsed after standing for 30 minutes; (3) Staining was performed with diluted LipidTOX Red staining solution for 30 minutes in the dark; and (4) The sample was fixed onto a slide, observed, and photographed using a confocal laser scanning microscope (FV1200; Olympus, Tokyo, Japan).

Thin layer chromatography

(1) The well-grown bacterial solution (4-8 mL) was collected in a centrifuge tube and centrifuged at 3000 g for 15 minutes to collect the bacteria; (2) The supernatant was removed, and the lipid was extracted using a mixture of PBS/MeOH/CHCl₃ (1:1:2, v/v/v). The bacterial solution was thoroughly vortexed three times; (3) The vortexed sample was centrifuged at 15000 rpm for 10 minutes at 4 ºC. The resulting separated phases are shown in Supplementary Figure 2; (4) A long, slender pipette tip was used to reach the bottom of the centrifuge tube to aspirate the lower lipid phase into a new EP tube. CHCl₃ was added to the tube for a second extraction; (5) The lipids obtained from the two extractions were dried and collected using high-purity nitrogen gas; (6) The obtained lipids were dissolved in an appropriate amount of CHCl₃ and spotted onto a silica gel plate; (7) The lipids were separated using thin layer chromatography (TLC) in a developing solvent of hexane/(C₂H₅)₂O/AA (80:20:1, v/v/v); and (8) After iodine staining, the target lipids on the plate were scraped into an EP tube and identified using gas chromatography-MS (Agilent 7000C Quadruple MS/MS with 7890B GC System; Agilent Technologies, Santa Clara, CA, United States). The analysis was conducted using the SCI-go research platform.

Purification of LDs

LDs were isolated with a modified method[17,21]: (1) Large-scale culture was performed, followed by centrifugation at 4000 g for 30 minutes. The bacterial pellet was collected and washed with buffer A; (2) An appropriate amount of buffer A and buffer B was prepared, stirred until dissolved, and stored at 4 ºC for later use; (3) The bacterial pellet, with the supernatant removed, was treated with buffer A (approximately 10 mL), and vortexed thoroughly to disperse the bacterial cells; (4) Before cell disruption, the bacterial solution (10 mL) was treated with phenylmethanesulfonyl fluoride (10 μL; 1000 ×). The bacterial cells were crushed at 1000-1500 bar using a low-temperature ultra-high cell crusher, with crushing repeated 3-5 times (if the cells are difficult to disrupt, lysozyme treatment was performed beforehand); (5) The disrupted bacterial solution was centrifuged at 6000 g for 10 minutes at 4 °C, and the supernatant was transferred to a new tube, forming the whole cell lysate; (6) After reserving a sample, the remaining lysate was divided equally into ultracentrifuge tubes (SW40; Beckman Coulter, Brea, CA, United States). Buffer B was gently layered on top of the liquid; (7) The tubes were precisely balanced and then placed in an ultracentrifuge (SW40) for centrifugation at 38000 rpm for 2 hours at 4 ºC; and (8) After ultracentrifugation, the SW40 tubes were gently removed and placed on ice. The top layer contained the LD sample. The LD phase, if clearly separated, was extracted or placed into an EP tube using a pipette tip rinsed with buffer B. The middle layer contained was cytosol, and the bottom layer represented the total membrane.

In vivo evaluation of A. pasteurianus lipids’ hypoglycemic effects

Establishment of diabetic mouse models: Preparation of citrate buffer: CA (2.10 g) was added to double-distilled water (100 mL) to create a stock solution, referred to as solution A. Trisodium citrate (2.94 g) was added to double-distilled water (100 mL) to create a stock solution, referred to as solution B. Solutions A and B were mixed in a 1:1.32 ratio, and the pH was adjusted to 4.2-4.5, resulting in a sodium citrate buffer solution (0.1 mol/L).

Animal selection: 42 SPF C57BL/6J female mice, aged 6 weeks and weighing 20 ± 2 g, were selected and acclimated for 5 days with free access to food and water. Six mice were randomly selected as the normal control group, while the remaining mice were designated for modeling.

Establishment of mice model treated with STZ: STZ was prepared at a concentration of 5 mg/mL in sodium citrate buffer (0.1 mol/L), protected from light and kept on ice. Mice were fasted for 10 hours and administered with STZ at a dose of 0.15 mL/10 g (75 mg/kg) via intraperitoneal injection for 5 days, consecutively. Following each injection, the mice remained fasted (with free access to water) for 90 minutes. On the 7^th day after the final injection, the mice were fasted (with water access) for 10 hours before blood collection, with their fasting blood glucose (FBG) measured from tail vein blood using a Roche glucometer (Roche, Indianopolis, IN, United States). Mice with FBG ≥ 16.7 mmol/L suggested an established diabetic mice model.

Grouping and administration: The diabetic mice were randomly divided into the following groups: Model group (PBS group), positive control group treated with Met (320 mg/kg), coenzyme Q9 (CoQ9) group (high concentration 75 mg/kg, and low concentration 25 mg/kg), and oleic acid (OA) group (high concentration 1000 mg/kg, and low concentration 500 mg/kg). Each group included six mice, each receiving 0.6 mL by gavage.

In vivo evaluation of therapeutic effects: After successful model establishment, administration by gavage was performed once daily, starting on the 2^nd day, and continued consecutively for 15 days. The normal control group and model group were given an equal volume of PBS, while the positive control group and the CoQ9 and OA high- and low-concentration groups received the corresponding treatments.

Blood glucose and body weight monitoring, along with general observations: Beginning on the first day of treatment, FBG and body weight were measured every 3 days. Additionally, the general condition of the mice was observed, including activity level, alertness, food and water intake, urination and defecation, and the moisture level of bedding. After 15 days of treatment, the mice were fasted for 10 hours. Blood was collected from the eyeball and centrifuged at 3000 g for 10 minutes to obtain serum. The levels of urea, creatinine, and uric acid were measured using the urease-glutamate dehydrogenase method, enzymatic method, and colorimetric method, respectively. The mice were sacrificed with tissues taken from their liver, spleen, and kidneys for hematoxylin and eosin (H&E) staining to observe tissue injury and repair. The pancreas was also collected, with half of the tissue fixed in formaldehyde for sectioning and staining to examine the effects of CoQ9 and OA on islets using a microscope. The other half was placed in the electron microscope (EM) fixative for ultrastructural observation of the islets after sectioning under an EM. The apoptosis-related B-cell lymphoma 2-associated X gene was detected using immunohistochemistry (terminal deoxynucleotidyl transferase dUTP nick end labeling [TUNEL] staining), allowing for the observation of cell apoptosis. The histology sectioning was performed by Wuhan Servicebio Technology Co., Ltd. (Hubei, China), and EM analysis was conducted by Shanghai Micronano Testing Technology Group Co., Ltd. (Shanghai, China).

Statistical analyses

All obtained data are presented as the mean ± SD, and statistical analyses were processed using SPSS 26.0 software (IBM SPSS Statistics, Armonk, NY, United States). Statistical analyses were conducted using either a t test for two groups of comparisons or an analysis of variance for multiple group comparisons. Graphs were generated via GraphPad Prism 8 (GraphPad Software, Boston, MA, United States). P < 0.05 was considered statistically significant. The smaller the P value, the greater the difference between the groups being compared.

RESULTS

Lipid metabolism of A. pasteurianus

Reliability and stability of the metabolic model: Principal component analysis was performed on the samples to preliminarily understand the overall metabolic differences between groups and the variability within each group. This analysis indicated the separation trends of lipid components among the groups and assessed whether there were significant differences between them. A principal component analysis model was established using LC-MS data to identify the overall differences among A. pasteurianus groups cultured in different media: BHI group, MSM group, and chromium- and zinc-enriched group (Supplementary Figure 3). The results showed that samples within each group clustered closely, while samples between groups were well separated, indicating good experimental reproducibility and differences between groups. The difference between the chromium- and zinc-enriched group and the BHI group was more pronounced than that between the MSM group and the BHI group. In addition, to improve sensitivity and separation, metabolic differences between groups were identified using partial least squares discriminant analysis and orthogonal partial least squares discriminant analysis (results are shown in Supplementary Figures 2 and 4). These analyses also suggested a stable and reliable metabolic model.

Differential analysis of lipids: Differential analysis was performed on all detected lipid molecules, with P < 0.05 and variable importance in the projection > 1 being the criteria for selection. As shown in Supplementary Table 1, compared to A. pasteurianus cultured in BHI, those cultured in MSM exhibited 122 downregulated differential lipids, with sphingolipids (SPs) being the most numerous categories. However, there were only four upregulated differential lipids; two of these were glycerophospholipids, and the other two were a fatty acyl and an SP, respectively. After the induction of culture by chromium and zinc enrichment, 258 differential lipids were downregulated, with SPs being the most numerous category. There were 243 upregulated differential lipids, with glycerolipids being the most numerous categories. The lipid profiles and differential lipids for each comparison group are illustrated in the volcano plots and scatter plots shown in Supplementary Figures 5 and 6, respectively. The cluster heatmap (Figure 1) indicated good clustering within groups and differences between groups. Metabolic pathway analysis was conducted to compare the differences in metabolic pathways of A. pasteurianus cultured in different media. The Kyoto Encyclopedia of Genes and Genomes annotation results indicated (Figures 2 and 3) that differences of functions of the differential lipids between the MSM group and the BHI group were primarily on SP metabolism, fatty acid biosynthesis, metabolic pathways, and glycerolipid metabolism. The functions of the differential lipids between the FF group (i.e. rich in chromium and zinc group) and the BHI group mainly concentrated in glycerolipid metabolism, metabolic pathways, SP metabolism, teichoic acid biosynthesis, inositol phosphate metabolism, and the metabolism of glycine, serine, and threonine. Under different culture conditions, A. pasteurianus adjusted these pathways to alter the types and quantities of lipids produced.

Figure 1 Cluster heatmap of differential lipids between groups. A: Mineral salt medium (MSM) vs brain heart infusion (BHI); B: FF (i.e. rich in chromium and zinc group) vs BHI. Metabolomics testing was performed on Acetobacter pasteurianus from BHI culture, chromium-rich zinc-rich culture, and MSM culture. After normalization of different contents, red indicates high content and green indicates low content. In the figure, the cluster lines on the left represent lipid clustering, and those on the top represent sample clustering. AP: Alkylphospholipid; BA: Bile acid; CAR: Cardiolipin; CASE: Sitosterol acetate; Cer-AP: Ceramide-1-phosphate; Cer-AS: Ceramide-1-sulfate; Cer-NDS: Ceramide-N-phosphoethanolamine; Cer-NP: Ceramide-N-phosphocholine; CoQ: Coenzyme Q; DG: Diacylglycerol; DGDG: Digalactosyldiacyl glycerol; FFA: Free fatty acid; LPA: Lysophosphatidic acid; LPC-O: Lysophosphatidyl choline-plasmalogen; LPE: Lysophosphatidyl ethanolamine; LPE-P: Lysophosphatidyl ethanolamine-plasmalogen; LPG: Lysophosphatidyl glycerol; LPI: Lysophosphatidyl inositol; MG: Monoacylglycerol; MGDG: Monogalactosyldiacylglycerol; NS: N-sphingomyelin; PA: Phosphatidic acid; PC: Phosphatidylcholine; PC-O: Plasmalogen phosphatidylcholine; PE-O: Plasmalogen phosphatidylethanolamine; PE-P: Phosphatidylethanolamine-plasmalogen; PG: Phosphatidylglycerol; PI: Phosphatidylinositol; PMeOH: Phosphatidyl methanol; PS: Phosphatidylserine; SM: Sphingomyelin; SPH: Sphingosine; TG: Triacylglycerol; TG-O: Triacylglycerol-plasmalogen.

Figure 2 Kyoto Encyclopedia of Genes and Genomes classification chart of differential lipids. A: Mineral salt medium (MSM) vs brain heart infusion (BHI); B: FF (i.e. rich in chromium and zinc group) vs BHI. Metabolomics testing was performed on Acetobacter pasteurianus from BHI culture, chromium-rich zinc-rich culture, and MSM culture.

Figure 3 Kyoto Encyclopedia of Genes and Genomes enrichment chart of differential lipids. A: Mineral salt medium (MSM) vs brain heart infusion (BHI); B: FF (i.e. rich in chromium and zinc group) vs BHI. Metabolomics testing was performed on Acetobacter pasteurianus from BHI culture, chromium-rich zinc-rich culture, and MSM culture.

Main lipids in the metabolism of A. pasteurianus

Qualitative and quantitative analyses of the lipids contained in A. pasteurianus were conducted. It was found that A. pasteurianus cultured in BHI medium had the highest overall lipid content (Supplementary Table 2, Figure 4A), and the three lipids with the highest contents were CoQ, free fatty acids (FFAs), and phosphatidylcholine (Figure 4B). From the most abundant lipid subclasses, CoQ and FFAs, the lipid with the highest content in CoQ was CoQ9 (Figure 4C), and in FFAs, the most abundant lipid was OA (FFA 18:1) (Figure 4D).

Figure 4 Lipid analysis of Acetobacter pasteurianus. A: Changes in the content of lipid subclasses; B: Total lipid content variation; C: Comparison of the contents of three types of coenzyme Q; D: Comparison of the contents of 24 types of free fatty acids. Metabolomics testing was performed on Acetobacter pasteurianus from brain heart infusion (BHI) culture, chromium-rich zinc-rich culture, and mineral salt medium culture. AP: Alkylphospholipid; BA: Bile acid; CAR: Cardiolipin; CASE: Sitosterol acetate; CE: Cholesteryl ester; Cer-ADS: Ceramide-1-acetate; Cer-AP: Ceramide-1-phosphate; Cer-AS: Ceramide-1-sulfate; Cer-NP: Ceramide-N-phosphocholine; Cer-NDS: Ceramide-N-phosphoethanolamine; Cer-NS: Ceramide-N-sulfate; CoQ: Coenzyme Q; DG: Diacylglycerol; DGDG: Digalactosyldiacyl glycerol; DG-O: Diacylglycerol-plasmalogen; FFA: Free fatty acid; LPA: Lysophosphatidic acid; LPC: Lysophosphatidyl choline; LPC-O: Lysophosphatidyl choline-plasmalogen; LPE: Lysophosphatidyl ethanolamine; LPE-P: Lysophosphatidyl ethanolamine-plasmalogen; LPG: Lysophosphatidyl glycerol; LPI: Lysophosphatidyl inositol; LPS: Lipopolysaccharide; MG: Monoacylglycerol; MGDG: Monogalactosyldiacylglycerol; MSM: Mineral salt medium; PA: Phosphatidic acid; PC: Phosphatidylcholine; PE: Phosphatidylethanolamine; PE-O: Plasmalogen phosphatidylethanolamine; PE-P: Phosphatidylethanolamine-plasmalogen; PG: Phosphatidylglycerol; PI: Phosphatidylinositol; PMeOH: Phosphatidyl methanol; PS: Phosphatidylserine; SM: Sphingomyelin; SPH: Sphingosine; TG: Triacylglycerol; TG-O: Triacylglycerol-plasmalogen.

LDs in A. pasteurianus metabolism

Limited presence and non-enrichment of neutral lipids in A. pasteurianus under standard culture conditions: After 48 hours of culture in an A. pasteurianus medium, neutral lipid staining revealed that some bacteria might contain neutral lipids. Lipid extraction and iodine staining were also performed, but no typical neutral lipid components of LDs were found (Figure 5A and B). Due to the long growth cycle and suboptimal growth conditions of A. pasteurianus in this medium, the composition of the medium was adjusted according to the requirements for bacterial growth and LD formation. Increasing the carbon source, specifically glucose content, revealed no change in the TLC results (Figure 5B). In combination with other existing cultivation methods[22,23], using BHI containing 2% glucose, peptone and other ingredients improved the growth condition of the strains; under the microscope, the bacterial cells appeared fuller, with some transitioning from short rods to long rods, which was favorable for the detection of LDs. However, despite lipid extraction and separation, no enrichment of neutral lipid components was observed (Figure 5C and D).

Figure 5 Detection of neutral lipids in Acetobacter pasteurianus under different nutritional conditions. A: Acetobacter pasteurianus cells were stained using neutral lipid dye LipidTOX red (100 ×); B: Thin layer chromatography (TLC) results of A. pasteurianus in media with different glucose concentrations; C: Comparison of the state and staining of A. pasteurianus before and after medium adjustment (100 ×); D: TLC results of A. pasteurianus after medium adjustment. DIC: Differential interference contrast; MLA: Monoacylglycerol; TAG: Triacylglycerol; WE: Wax ester.

However, the LDs of A. pasteurianus have not yet been successfully purified. The reasons for this may include: (1) A. pasteurianus has a very low LD content; (2) With the bacterial membrane difficult to disrupt, the LDs were not fully released; (3) LDs, often located near the cell membrane, were separated with the cell membrane components at the bottom of the tube after cell disruption and ultracentrifugation; and (4) The density and structure of the primary neutral lipid component, WE, differing from those of TAG, may not be separated to the top layer. Both WE and TAG are neutral lipid components of LDs, serving as storage compounds for energy and carbon, or as deposits of toxic or unused fatty acids. Apart from differences in density and structure, the shapes of these lipids in cells also vary. TAG is generally enclosed within spherical lipid bodies and is the most common core lipid component extracted from LDs, while WE are not confined to spherical shapes and may also exist in flattened, disc-like, or rectangular forms[18].

Enrichment of WE in A. pasteurianus stimulated by low nitrogen, high carbon environment: Most bacterial LDs can be produced in large quantities under stress conditions such as low nitrogen, high carbon environments (high carbon/nitrogen ratio) or during unbalanced growth. All of these lipids serve as storage compounds for energy and carbon, maintaining metabolism during starvation periods, which may be advantageous for their survival and evolution under adverse conditions[24]. Placing a large number of well-growing strains into low nitrogen MSM resulted in an increase in neutral lipids in A. pasteurianus (Figure 6A). Lipid extraction and TLC analysis revealed an enrichment of WE (Figure 6B). The target lipids were identified to contain WE (belonging to long-chain fatty acid esters) and other SCFA esters, as shown in Supplementary Figures 7 and 8. A-H are sequentially 2-butenyl butyrate, butyl butyrate, decyl formate, methyl butyrate, 10-methyl undecanoate methyl ester, isobutyl butyrate, octyl stearate, and 2,6,10,14-tetramethylpentadecanoic acid methyl ester. WE, as one of the core components of LDs, are capable of storing energy. Naturally extracted or synthetically produced WE are widely used in food, cosmetics, and pharmaceutical products. A small intake in humans may have certain anti-inflammatory and anti-obesity effects[25-27], but they should not be consumed in excess.

Figure 6 Detection of neutral lipids in Acetobacter pasteurianus cultured in mineral salt medium. A: Acetobacter pasteurianus observed using neutral lipid staining (100 ×, spherical fluorescent sites indicated by arrows may contain lipid droplets); B: Thin layer chromatography results for the lipids of A. pasteurianus. DIC: Differential interference contrast; MLA: Monoacylglycerol; TAG: Triacylglycerol; WE: Wax ester.

LDs in A. pasteurianus

Random culture time points were selected to measure the optical density: 600 nm of A. pasteurianus in the original medium at different growth times, and a growth curve was established (Figure 7A). From 0-20 hours, A. pasteurianus was in the lag phase, 20-28 hours in the logarithmic growth phase, 36-48 hours in the stationary phase, and after 48 hours in the decline phase. A. pasteurianus cells collected during the stationary phase were cultured in MSM. Equal amounts of bacterial samples were taken at different time points for lipid extraction and TLC analysis, showing no significant differences in the contents of WE across the time points (Figure 7B). A. pasteurianus cells were collected after 48 hours of culture in MSM for LDs purification. A. pasteurianus LDs could not be purified using ultracentrifugation included in LDs extraction methods (typically including TAG).

Figure 7 Detection of lipid droplets in Acetobacter pasteurianus. A: Growth curve; B: Lipid analysis at various time points during culture in mineral salt medium. OD: Optical density; TAG: Triacylglycerol; WE: Wax ester.

Hypoglycemic effects of the main lipids of A. pasteurianus

Previously, we found that OA blocks palmitic acid-induced insulin resistance[28]. We speculated A. pasteurianus-produced OA may play similar role. In the diabetic mice with FBG ≥ 11.1 mmol/L, groups were given treatments for 15 days, with FBG measured every 3 days. The results showed that the high-concentration OA group experienced a significant decrease in blood glucose levels, better than those treated with Met. The high-concentration CoQ9 group and the low-concentration OA group had effects comparable to the Met group, while the low-concentration CoQ9 group showed slightly inferior results (Figure 8A and B). During this process, except for the OA group, other treatment groups maintained a good mental state, exhibited normal activity levels, and had normal drinking, eating, and excretion volumes, with the frequency of bedding moisture decreasing as the treatment duration increased. The CoQ9 and Met groups showed good weight recovery, while the OA group experienced significant weight loss (Figure 8C and D).

Figure 8 Changes in blood glucose and body weight during medication administration in diabetic mice. A: Diabetic mouse treatment model; B: Bar chart of changes in blood glucose levels; C: Line graph of changes in blood glucose levels; D: Chart of changes in body weight. ^nsP > 0.05. ^aP < 0.05. ^bP < 0.01. ^cP < 0.001. CoQ9: Coenzyme Q9; FBG: Fasting blood glucose; Met: Metformin; OA: Oleic acid; PBS: Phosphate-buffered saline; pH: Potential of hydrogen; STZ: Streptozotocin.

Repair effects on tissue of the main lipids of A. pasteurianus

After treatment in diabetic mice, the pancreatic tissue was treated using H&E staining and observed under a microscope. The results showed that the islets in the PBS group mice were reduced in number, with evident islet shrinkage and damage. In contrast, mice in the Met, CoQ9, and OA groups exhibited less islet shrinkage. The islet tissue recovery in the CoQ9 and OA groups was better than the Met group, with no inflammation observed around the islets. The pancreatic tissue from PBS group mice observed using transmission electron microscopy showed a reduction in secretory granules, mitochondrial swelling, cristae rupture, and ER dilation. After treatment with CoQ9 and OA, both the mitochondria and ER nearly returned to normal, with some recovery of damage, smaller ER dilation, and slight swelling mitochondria. Immunohistochemical results indicated a reduction in islet cell apoptosis in the CoQ9 and OA groups (Figure 9A), with effects comparable to those observed in the Met group. In addition, diabetes caused the kidneys to work overload, increasing blood flow and capillary pressure, leading to glomerular lesions. In the early stages, glomerular hypertrophy, homogeneous thickening of the glomerular basement membrane, and mesangial matrix proliferation were observed. The CoQ9 and OA groups showed more significant improvement in kidney functions, with near-normal recovery after treatment (Figure 9B). However, there were no significant differences in the serum indicators of renal function (Supplementary Table 3), possibly because the kidney damage in early-stage diabetic mice was not severe enough to fully reflect the repair effects. H&E staining of other organ sections showed no damage to the liver and spleen in the CoQ9 and OA groups, suggesting that they were safe and non-toxic (Supplementary Figure 9).

Figure 9 Role of Acetobacter pasteurianus lipids in promoting pancreatic and renal tissue repair. A: Images of pancreatic tissue observed using hematoxylin and eosin (H&E) staining, terminal deoxynucleotidyl transferase dUTP nick end labeling (TUNEL) staining, and electron microscopy; B: Images of renal tissue observed using H&E staining and TUNEL staining. The green arrows indicate the thickening of the glomerular basement membrane. CoQ9: Coenzyme Q9; Met: Metformin; OA: Oleic acid; PBS: Phosphate-buffered saline; TEM: Transmission electron microscopy.

DISCUSSION

BHI has good cultivation effects on a variety of microorganisms and is commonly used as a basic medium. Its nutritional components mainly come from peptone, glucose, and saline solutions. However, when cultured in BHI with 2% glucose, a high-nutrient medium, A. pasteurianus exhibited the highest lipid content compared to MSM and chromium- and zinc-rich media. A total of 583 lipid metabolic products were identified, with the secretion of significant amounts of lipid components such as CoQ9 and OA. When chromium and zinc were added to the culture, the lipid content of A. pasteurianus decreased, but the number of upregulated lipid species was the highest. CoQ9, a ubiquinone containing nine isoprenoid units, is a normal component of human plasma and shares the same cardioprotective effects as ubiquinone (CoQ10). OA, also known as cis-9-octadecenoic acid, is a naturally occurring unsaturated fatty acid that plays an important role in softening blood vessels and metabolism. CoQ has certain effects in controlling hyperglycemia and restoring insulin secretion[29], and it also improves mitochondrial dysfunction in the kidneys[30]. OA, a monounsaturated fatty acid, possesses anti-inflammatory properties, inhibits ER stress, prevents the attenuation of insulin signaling pathways, and enhances β-cell survival[31]. In this study, we confirmed that the unsaturated fatty acids CoQ9 and OA derived from A. pasteurianus have hypoglycemic effects and promote tissue repair. For the analysis of biochemical pathways that may be involved, CoQ is an important component of the mitochondrial electron transport chain and is involved in electron transport and adenosine triphosphate synthesis, and an increase in CoQ may disrupt the electron transport chain function, activate the adenosine monophosphate-activated protein kinase signaling pathway, promote glucose uptake and fatty acid oxidation, and inhibit gluconeogenesis, and also can ameliorate the effects of high glucose-induced apoptosis and dysfunction[32]. OA can be activated through β-oxidation, which can enter the tricarboxylic acid cycle to provide energy to the cells[33,34], and can indirectly affect glucose metabolism by reducing glucose dependence, thus lowering blood glucose levels, and enhancing insulin sensitivity through the phosphatidylinositol 3-kinase/protein kinase B signaling pathway[35,36]. So, there is good theoretical support for the molecular mechanisms of CoQ9 and OA on glucose metabolism.

In addition, neutral lipids are stored in a membrane-bound organelle[37], known as the LD, which is conserved across eukaryotic cells, prokaryotic bacteria, and ancient species[18]. Due to its complex surface proteins[38-40] which may influence human health and the development of metabolic diseases, its existence and related research has been a focus of attention. Thus, it is evident that LDs extracted from both tissue cells and bacteria can be studied from the perspectives of lipid metabolism and protein function for exploring their potential benefits for human health. This study primarily focused on the detection and extraction of lipid components and LDs from the probiotic A. pasteurianus. The results revealed that under low nitrogen, high carbon conditions, the major lipid components contained by A. pasteurianus were fatty acid esters, including WE. This suggests that A. pasteurianus may regulate the fatty acid biosynthesis pathway to promote the production of fatty acyl lipids, which are subsequently synthesized. However, the LDs of A. pasteurianus have not yet been successfully purified. The reasons for this may include: (1) A. pasteurianus has a very low LD content; (2) With the bacterial membrane difficult to disrupt, the LDs were not fully released; (3) LDs, often located near the cell membrane, were separated with the cell membrane components at the bottom of the tube after cell disruption and ultracentrifugation; and (4) The density and structure of the primary neutral lipid component, WE, differing from those of TAG, may not be separated to the top layer. Both WE and TAG are neutral lipid components of LDs, serving as storage compounds for energy and carbon, or as deposits of toxic or unused fatty acids. Apart from differences in density and structure, the shapes of these lipids in cells also vary. TAG is generally enclosed within spherical lipid bodies and is the most common core lipid component extracted from LDs, while WE are not confined to spherical shapes and may also exist in flattened, disc-like, or rectangular forms[18]. Whether it is necessary to purify LDs primarily composed of WE and how to do so remains to be further explored.

CONCLUSION

In conclusion, A. pasteurianus contains various lipid components, with the main components, CoQ9 and OA exhibiting hypoglycemic effects and promoting the repair of pancreatic and renal tissues. The composition also includes some neutral lipids, but LDs could not be purified. The lipid components of probiotics have good prospects for clinical applications and are of importance for treating metabolic diseases and improving human health. This study provides an important theoretical basis for the clinical application of A. pasteurianus as a probiotic.